Dialdehydes lead to exceptionally fast bioconjugations at neutral pH by virtue of a cyclic intermediate

Article abstract of DOI:10.1002/anie.201406132

One of the open challenges in chemical biology is to identify reactions that proceed with large rate constants at neutral pH values. As shown here, dialdehydes react with O-alkylhydroxylamines at rates of 500 M^-1 s^-1 at neutral pH va

Full text of DOI:10.1002/anie.201406132

Angewandte
Chemie
DOI: 10.1002/anie.201406132
Bioconjugation
Hot Paper
Dialdehydes Lead to Exceptionally Fast Bioconjugations at Neutral pH
by Virtue of a Cyclic Intermediate**
Pascal Schmidt, Linna Zhou, Kiril Tishinov, Kaspar Zimmermann, and Dennis Gillingham*
Abstract: One of the open challenges in chemical biology is to
identify reactions that proceed with large rate constants at
neutral pH values. As shown here, dialdehydes react with
O-alkylhydroxylamines at rates of 500mÀ1 sÀ1 at neutral
pH values in the absence of catalysts. The key to these
conjugations is an unusually stable cyclic intermediate, which
ultimately undergoes dehydration to yield an oxime. The scope
and limitations of the method are outlined, as well as its
application in bioconjugation and a mechanistic interpretation
that will facilitate further developments of reactions with
alkylhydroxylamines at low substrate concentrations.
C
ondensations of hydroxylamine and hydrazine derivatives
with carbonyl groups are widely used methods for bioconju-
gation.[1] At neutral pH values, however, oxime condensations
are too slow to be employed at the high dilution that is often
required in biological settings. The fastest small-molecule
bioconjugations are the tetrazine inverse-electron-demand
Diels–Alder reactions with strained olefins (k ꢀ 102–
104 mÀ1 sÀ1),[2] and a recently described amide-forming ligation
of acyltrifluoroborates with O-carbamoylhydroxylamines
(k = 20mÀ1 sÀ1).[3] We describe here that dialdehydes react
with O-alkylhydroxylamines with second-order rate constants
in the range of 102–103 mÀ1 sÀ1 at neutral pH values, thus
competing with the fastest small-molecule systems.[4] In
practical terms, rate constants of this magnitude mean that
even at concentrations of 5 mm for each substrate, the first
half-life is less than seven minutes. A mechanistic analysis
indicates that the key to these reactions is the rapid and
irreversible formation of an isoindole bis(hemiaminal)
(IBHA), which gradually dehydrates to the oxime (see
Figure 1c).
Figure 1. a) The Jencks mechanism for condensations with O-alkylhy-
droxylamines. b) Known approaches for accelerating the reaction are
shown in blue on the reaction coordinate diagram. c) The approach
outlined here is given in red: O-alkylhydroxylamines react with ortho-
phthalaldehyde (OPA) to give stable isoindole bis(hemiaminals)
(IBHAs) within seconds at neutral pH values (conditions: 10 mm
phosphate buffer, pH 7.1, 20% acetonitrile). Please note that the
reaction coordinate diagram is qualitative; smooth curves are drawn
between kinetically relevant steps although proton transfer steps and
intermediate addition products are likely to be involved.
Jencksꢀ seminal mechanistic work showed that oxime
condensations typically proceed through a rapid pre-equilib-
rium with a tetrahedral hemiaminal intermediate followed by
a rate-limiting dehydration (Figure 1a).[5] Most attempts to
optimize oxime and hydrazone condensations have therefore
focused on lowering the barrier of the dehydration step: The
use of low pH values,[6] nucleophilic catalysis,[7] and intra-
molecular proton assistance[8] are prominent examples (Fig-
ure 1b). At high dilution, however, the initial bimolecular
addition step becomes rate-limiting and therefore warrants
scrutiny. Herein, we show that dialdehydes provide an
internal trap to stabilize hemiaminals as IBHAs, leading to
an enormous shift in the equilibrium position. The Keq is so
large that the initial addition step becomes effectively
irreversible; this simplifies hydroxylamine-based bioconjuga-
tions because the timing of the typically slow dehydration
reaction is rendered immaterial: Both IBHAs and oximes
represent successful conjugations. The results we present
should find broad applicability in the labelling of biomole-
cules at high dilution and provide a framework for engineer-
ing further variants of O-alkylhydroxylamine addition reac-
tions.
[*] P. Schmidt,[+] K. Tishinov, K. Zimmermann, Prof. Dr. D. Gillingham
Department of Chemistry, University of Basel
St. Johanns-Ring 19, 4056 Basel (Switzerland)
E-mail: dennis.gillingham@unibas.ch
Dr. L. Zhou[+]
Chemistry Research Laboratory, University of Oxford
12 Mansfield Road, Oxford, OX1 3TA (UK)
[+] These authors contributed equally to this work.
[**] This work was supported by start-up funds from the University of
Basel and a Novartis postgraduate fellowship to L.Z.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201406132.
Angew. Chem. Int. Ed. 2014, 53, 1 – 5
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1
These are not the final page numbers!

.
Angewandte
Communications
Table 1: Comparison of oxime ligation reactions with different aldehydes and
While testing the propensity of various ortho-
substituted aldehydes to accelerate oxime conden-
sations, we discovered that a stable intermediate
formed within seconds of mixing ortho-phthalalde-
hyde (OPA) with O-benzylhydroxylamine in a 1:1
ratio at 100 mm at neutral pH (Figure 1c). A
complete NMR characterization of the intermediate
suggested a 9:1 ratio of diastereomers of the IBHA
structure that is shown in red in Figure 1.[9] The
IBHA further dehydrated to give the expected
oxime over a time scale of hours. The rate of
intermediate formation and its stability under
neutral conditions immediately suggested its poten-
tial in bioconjugation. Indeed morpholino-type
bis(hemiaminals) have been observed during the
3’-end labelling of periodate-cleaved RNAs;[10]
surprisingly, however, RNA 3’-labelling has never
been subjected to a detailed kinetic characteriza-
tion, nor has it inspired a general bioconjugation
approach based on dialdehydes.
Two reports have described accelerating effects
of neighboring substituents on oxime or hydrazone
condensations: Jencks and Wolfenden observed up
to a tenfold rate enhancement with certain ortho-
substituted benzaldehydes,[11] an effect that they
attributed to a change in the addition equilibrium
owing to destabilization of the starting aldehyde
(indicated in blue in Figure 1). Kool et al. have
recently shown that neighboring protons facilitate
hydrazone condensations, and they proposed that
intramolecular proton assistance helped the dehy-
dration (blue annotation in Figure 1).[8a] The stabi-
lization of adducts prior to dehydration, such as with
IBHAs, would offer a new way of controlling
alkylhydroxylamine conjugations.
hydroxylamines.[a]
Entry
1[b]
Aldehyde
RONH2
Product
Conv. [%]
<1
2[b]
3[c]
<10
>98
4[c]
>98
5[b]
6[b]
<10
<10
7[b]
<1
8[b]
9[b]
98
79
[a] Note that the oxime products are reported as acidic LC-MS detection was
employed; with OPA, the adducts present prior to LC-MS injection are mixtures of
the IBHAs and oximes. Each reaction in Table 1 that employs ortho-dialdehydes
leads to complete conversion, we show the 90 min time point for convenient
comparison. [b] Substrate concentrations: 100 mm; phosphate buffer (100 mm),
pH 7.2. [c] Substrate concentrations: 10 mm.
To explore the nature of the acceleration and to
establish the reaction scope, we first varied the
substrates (Table 1). A survey of carbonyl derivatives con-
firmed the primacy of OPA: The reactions of benzaldehyde
(entry 1), 2-formylbenzoic acid (entry 2), diacetylbenzene
(entry 7), and dialdehydes with other substitution patterns
(entries 5 and 6) were significantly slower. As shown in
entry 4, bis(oximes) can also be formed, an approach that may
be important when the presence of a residual aldehyde is
undesirable. The special reactivity of dialdehydes is not
limited to OPA; for example naphthalene dialdehyde was also
a highly effective substrate, converting completely into the
corresponding oxime within minutes (entry 8). The acceler-
ation with OPA is maintained with complex substrates since
a pentapeptide (LYRAG) bearing an N-terminal hydroxyl-
amine gave 79% conversion after 1.5 hours at concentrations
of 100 mm in each reaction partner (entry 9).
Time course NMR analysis provided a value for k2 and
a lower bound for Keq (see Figure 2). As the dehydration is
effectively irreversible over the time scales we monitored,[12]
the appearance of oxime could be taken as a direct measure of
k2 (see the Hc protons in Figure 2). The value obtained, 1.2 Æ
0.2 ꢁ 10À5 sÀ1, is two to three orders of magnitude smaller than
those of typical oxime condensations at pH 7,[7a] consistent
with the intermediate IBHA sitting in a deep energy
minimum. No equilibrium for the initial addition was
measurable over the observed time scale (20 h), a point
independently confirmed in a separate experiment: When the
IBHA[9] was dissolved in phosphate buffer, and the reaction
mixture monitored by NMR spectroscopy, the formation of
OPA was not observed; instead, the substrate was gradually
converted into the oxime. Although an accurate Keq value
could not be determined, a lower bound of approximately
107 mÀ1 may be assumed since low micromolar concentrations
of the aldehyde would be well above the detection limit of the
700 MHz NMR spectrometer. In terms of the reaction
coordinate diagram (see Figure 1), the fact that the IBHA
led to oxime formation but not to formation of OPA suggests
that dehydration is faster than retro-IBHA formation (i.e.,
k2 > kÀ1).[13] Overall, the mechanism of oxime formation that
we observe with dialdehydes parallels Jencksꢀ proposal for
simple oximes (see Figure 1), with the key distinction that the
equilibrium constant for the formation of the intermediate
overwhelmingly favors addition.
The formation of the IBHA was so rapid that even at
a concentration of 100 mm, NMR analysis lacked the time
2
www.angewandte.org
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2014, 53, 1 – 5
These are not the final page numbers!

Angewandte
Chemie
and the instability of the tetrazines. In comparison, IBHA
formation has rate constants that compete with the fastest
cycloaddition reactions,[4] but employs far simpler substrates.
For the labelling of complex samples, the dialdehyde must
tolerate other biological nucleophiles. With simple amines,
dialdehydes can form a great variety of products depending
on the reaction conditions. Reactions with amines, however,
are slower than what we observe with alkylhydroxylamines,
and, although postulated as intermediates, IBHAs have never
been observed with amines under aqueous conditions.[16]
Under dilute conditions in cold DMSO, an ammonia-derived
IBHA has been observed by NMR spectroscopy, but it rapidly
eliminated water to deliver the Schiff base.[17] In studies using
dialdehydes to capture substrate–kinase pairs, the groups of
Shokat[18] and others[19] have shown that reactions with
amines[20] can be mitigated by employing less reactive
dialdehydes. In our case, the IBHA conjugation was equally
effective in the presence of equimolar tryptophan. In the
presence of lysozyme, however, the conversions were lower
(see entries 3 and 6, Table 2), likely as a result of nucleophilic
side-chains in the protein trapping some of the dialdehyde.
Table 2: Effect of amino acids and proteins on the IBHA reaction.[a]
Figure 2. Kinetic analysis. a) NMR analysis indicating the peaks that
were monitored with time to determine the value of k2. NMR spectra
covering the entire ppm range are included in the Supporting
Information, Figure S8. Kinetic parameters are an average of two
experiments. b) A fluorescence quenching assay was used to establish
the rate of the initial bimolecular addition. k1 is the combined average
of six repeats at two different concentrations (see Figure S13 for the
kinetic plots).
Entry Dialdehyde
Additive
Product
Conv. [%]
1
2
3
–
79
67
37
tryptophan
lysozyme
4
5
6
–
77
76
57
tryptophan
lysozyme
resolution to measure the initial bimolecular rate constant k1.
However, a value for k1 is important in the planning of
experiments because it forecasts the required concentrations
and stoichiometry. We therefore developed a fluorescence
quenching assay to determine k1. A lissamine-tagged version
of OPA and a dabcyl derivative bearing a hydroxylamine
were synthesized, and kinetic parameters were determined at
concentrations of 5 mm and 2 mm for each substrate (100 mm
phosphate buffer, pH 7.2). The resulting data fit best to
a second-order rate equation and gave an average over the
two concentrations of k1 = 494 Æ 108mÀ1 sÀ1 (see Figure 2b).
Although the fluorescence quenching assay was run with
different substrates (the tags necessary for the readout), the
magnitude of k1 is qualitatively consistent with the rapid
formation of the intermediate that was observed by NMR
spectroscopy. Using these numbers we can now annotate the
major elementary steps for oxime formation with dialdehydes
(bold-faced values in Figure 2). The most rapid extant small-
molecule bioconjugation is the tetrazine-strained olefin
Diels–Alder reaction, which typically has second-order rate
constants between 1 and 103 mÀ1 sÀ1,[2a,b,14] and in the best cases
up to 106 mÀ1 sÀ1.[4,14–15] Despite these impressive rate con-
stants, the widespread adoption of the tetrazine Diels–Alder
reaction has been hindered by challenging substrate syntheses
[a] Please note that we report the oxime products since acidic HPLC
analysis was employed; with OPA, the adducts present prior to HPLC
injection are mixtures of the IBHAs and oximes.
These results suggest that alkylhydroxylamine additions may
be conducted with complex samples, but that a slight excess of
the dialdehyde may be necessary for complete conversion.
To further demonstrate the practical potential of the
method we examined a bioconjugation with a complex DNA
substrate. We synthesized a 5’-O-alkylhydroxylamine termi-
nated DNA 41-mer and tested it with various concentrations
of an OPA derivative bearing the lissamine fluorophore
(Figure 3, top). As the gel data illustrate, under neutral
conditions, even down to 5 mm of DNA, almost complete
conversion into the labelled product was observed after one
hour (compare lanes 3–7), whereas the control lanes using
either a monoaldehyde (lane 8) or a native DNA (lane 9)
showed no labelling. Given the proclivity of nucleic acids to
depurinate at the low pH values needed for most oxime
condensations, the neutral conditions employed here provide
a mild alternative for DNA and RNA bioconjugations.[21]
Angew. Chem. Int. Ed. 2014, 53, 1 – 5
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3
These are not the final page numbers!

.
Angewandte
Communications
Keywords: aldehydes · bioconjugation · hydroxylamines ·
.
kinetics · oximes
[1] G. T. Hermanson, Bioconjugate techniques, Elsevier/Academic
Press, Amsterdam, 2008.
[2] a) N. Devaraj, R. Weissleder, Acc. Chem. Res. 2011, 44, 816 – 827;
b) M. L. Blackman, M. Royzen, J. M. Fox, J. Am. Chem. Soc.
2008, 130, 13518 – 13519; c) M. Taylor, M. Blackman, O.
Dmitrenko, J. Fox, J. Am. Chem. Soc. 2011, 133, 9646 – 9649.
˝
[3] H. Noda, G. Eros, J. W. Bode, J. Am. Chem. Soc. 2014, 136,
5611 – 5614.
[4] D. M. Patterson, L. A. Nazarova, J. A. Prescher, ACS Chem.
Biol. 2014, 9, 592 – 605.
[5] W. P. Jencks, Progress in Physical Organic Chemistry, Wiley,
Hoboken, 1964, pp. 63 – 128.
[6] E. H. Cordes, W. P. Jencks, J. Am. Chem. Soc. 1962, 84, 4319 –
4328.
[7] a) A. Dirksen, T. M. Hackeng, P. E. Dawson, Angew. Chem. Int.
Ed. 2006, 45, 7581 – 7584; Angew. Chem. 2006, 118, 7743 – 7746;
b) M. Rashidian, M. M. Mahmoodi, R. Shah, J. K. Dozier, C. R.
Wagner, M. D. Distefano, Bioconjugate Chem. 2013, 24, 333 –
342; c) P. Crisalli, E. T. Kool, J. Org. Chem. 2013, 78, 1184 – 1189.
[8] a) E. T. Kool, D.-H. Park, P. Crisalli, J. Am. Chem. Soc. 2013, 135,
17663 – 17666; b) P. Crisalli, E. T. Kool, Org. Lett. 2013, 15,
1646 – 1649.
[9] The IBHA could be preparatively isolated as a mixture with the
oxime product by performing the reaction in acetonitrile. See the
Supporting Information, Section 4.4, for full details of the
procedure and a full NMR characterization of the IBHA.
[10] a) J. A. Hunt, Biochem. J. 1965, 95, 541 – 551; b) F. Hansske, F.
Cramer, Methods Enzymol. 1979, 59, 172 – 181; c) F. Hansske, M.
Sprinzl, F. Cramer, Bioorg. Chem. 1974, 3, 367 – 376; d) D. K.
Willkomm, R. K. Hartmann in Handbook of RNA Biochemistry,
Wiley-VCH, Weinheim, 2008, pp. 86 – 94; e) J. Summerton, D.
Weller, Antisense Nucleic Acid Drug Dev. 1997, 7, 187 – 195.
[11] R. Wolfenden, W. P. Jencks, J. Am. Chem. Soc. 1961, 83, 2763 –
2768.
Figure 3. Top: DNA was efficiently labelled under mild conditions
(pH 7). Please note that for simplicity, the IBHA is drawn, but the
conjugation products were likely to be a mixture of IBHA and oxime.
Gel: Lane 1: lissamine dialdehyde. Lane 2: 41-mer 5’-O-alkylhydroxyl-
amine DNA. Lanes 3–7: IBHA conjugations at various DNA concen-
trations. Lane 8: control reaction of alkylhydroxylamine DNA with the
fluorescent monoaldehyde 2-naphthaldehyde. Lane 9: control reaction
confirming that the lissamine dialdehyde does not react with a native
DNA oligomer (39-mer). Note that the reactions were run in unbuf-
fered water; the pH value measured at the end of the reaction was 7.
[12] J. Kalia, R. T. Raines, Angew. Chem. Int. Ed. 2008, 47, 7523 –
7526; Angew. Chem. 2008, 120, 7633 – 7636.
[13] A separate NMR experiment also supports that k2 > kÀ1, and
that the first step is not an equilibrium: When a sample
containing the IBHA of napthalene dialdehyde was treated
with a large excess of OPA, no scrambling to the OPA-derived
IBHA was observed. This suggests that once the IBHA is
formed, the alkylhydroxylamine is never released again. See the
Supporting Information, Section 4.3, for the NMR experiments
supporting this conclusion.
Most effort in optimizing oxime and hydrazone conden-
sations has been devoted to accelerating the final dehydration
step as it has the highest activation barrier. However, at
sufficiently low concentrations, bimolecular elementary steps
will begin to dominate the rate equation of any reaction
irrespective of barrier heights. Further development of
alkylhydroxylamine bioconjugations at low concentrations
will therefore require an increased focus on the initial
bimolecular addition step. The propensity of OPA to form
a stable IBHA is one such example, but additional possibil-
ities could be imagined: Other dicarbonyl compounds that
lead to stable IBHAs or alkylhydroxylamines that also deliver
stable intermediates present new targets for future studies.
[14] See Ref. [2c].
[15] R. Selvaraj, J. M. Fox, Curr. Opin. Chem. Biol. 2013, 17, 753 –
760.
[16] a) E. Kulla, P. Zuman, Org. Biomol. Chem. 2008, 6, 3771 – 3780;
ˇ
b) J. Klꢂma, M. Polꢃsek, J. Ludvꢂk, J. Urban, J. Heterocycl. Chem.
2012, 49, 1202 – 1209.
[17] T. Do Minh, A. L. Johnson, J. E. Jones, P. P. Senise, J. Org. Chem.
1977, 42, 4217 – 4221.
[18] A. V. Statsuk, D. J. Maly, M. A. Seeliger, M. A. Fabian, W. H.
Biggs, D. J. Lockhart, P. P. Zarrinkar, J. Kuriyan, K. M. Shokat, J.
Am. Chem. Soc. 2008, 130, 17568 – 17574.
[19] K. Liu, K. A. Kalesh, L. Bing Ong, S. Q. Yao, ChemBioChem
2008, 9, 1883 – 1888.
[20] P. Zuman, Chem. Rev. 2004, 104, 3217 – 3238.
Received: June 11, 2014
[21] An
additional
example
of
DNA
labelling
with
Revised: July 18, 2014
Published online: && &&, &&&&
a bis(alkylhydroxylamine) is also provided in the Supporting
Information, Section 2.3.
4
www.angewandte.org
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2014, 53, 1 – 5
These are not the final page numbers!

Angewandte
Chemie
Communications
Bioconjugation
P. Schmidt, L. Zhou, K. Tishinov,
K. Zimmermann,
D. Gillingham*
&&&& — &&&&
Dialdehydes Lead to Exceptionally Fast
Bioconjugations at Neutral pH by Virtue
of a Cyclic Intermediate
For aldehydes, 1+1=500: Dialdehydes
react with O-alkylhydroxylamines at rates
of 500mÀ1 sÀ1 at neutral pH values in the
absence of catalysts. The key to these
conjugations is an unusually stable cyclic
intermediate, which ultimately undergoes
dehydration to yield an oxime. The appli-
cation of this method in bioconjugations
and a mechanistic interpretation are
outlined.
Angew. Chem. Int. Ed. 2014, 53, 1 – 5
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5
These are not the final page numbers!